Glioblastoma (GBM) is a fatal primary brain tumor, with most patients surviving 12-15 months following diagnosis, despite receiving standard of care treatment. We discovered that human cytomegalovirus (HCMV) is present in a high percentage of GBM, and these findings have been confirmed by multiple groups. Our co- investigator, Dr. Nino Chiocca, has recently demonstrated that CMV infection accelerates tumor progression to GBM in a transgenic mouse model of glioma. Recently, antiviral therapies aimed at HCMV have demonstrated efficacy in patients with GBM. We have determined that Cidofovir (CDV), an FDA approved antiviral agent for CMV infection, potently inhibits GBM proliferation both in vitro and in an intracranial xenograft mouse model of primary GBM. Furthermore, our investigations of CDV have indicated that, while the antiviral effect is significant to the antitumr response, CDV possesses potent antitumor properties even in the absence of CMV infection. Importantly, the antitumor response seen using the in vivo GBM xenograft model with CDV is greater than any systemic agent tested thus far at UCSF other than temozolomide, and for temozolomide, substantial tumor response is limited to the use of tumor cell sources and their corresponding xenografts having methylated MGMT promoter. Mechanistically, we have determined that CDV, a cytosine analog, incorporates not only into CMV DNA but also into tumor cell genomic DNA. We find that CDV uptake by tumor cells activates apoptotic pathways and blocks normal DNA repair mechanisms, which are responsible for tumor cell resistance to and recovery following radiation therapy (RT), the primary standard of care treatment modality for newly-diagnosed GBM. In our in vitro and in vivo studies, CDV enhances the effect of RT. Although poor tissue uptake and nephrotoxicity are limitations of using CDV in treating human disease, our collaborator Dr. Karl Hostetler, has synthesized CDV derivatives that are highly lipophilic, and can be administered orally while causing minimum toxicity. In our preliminary studies, these compounds exhibit > 300-fold increased antitumor effect in vitro, thus allowing dramatically lower dosages to be administered, in comparison with cidofovir, for achieving significant anti-tumor activity. In Aim 1, we plan to use existing mouse xenograft models developed by our co-investigator Dr. David James at UCSF to optimize treatments based on acquired knowledge of investigational agent pharmacokinetics and to determine optimal regimens to use when combining inhibitors with RT. In Aim 2 we will utilize Dr. Chiocca's transgenic mouse model of spontaneously occurring glooms +/- CMV infection to understand the impact of antiviral therapy in the setting of CMV-mediated GBM tumor genesis. In addition, we will examine the impact of antiviral therapy with respect to the influence of CMV on the tumor microenvironment and host immune response to tumor. Finally, in Aim 3 we plan to investigate the mechanisms by which CDV and CDV-derived compounds elicit DNA damage and interfere with DNA repair mechanisms. We anticipate that one of the major advantages of these compounds will be their high therapeutic index: i.e., limited toxicity in normal cells, but dramati apoptotic pathway activation and prevention of DNA repair in tumor cells, which could significantly delay or possibly prevent radiation and chemo-resistance critical for GBM recurrence.